simplest speaker crossover

Status
Not open for further replies.
Jump to Latest
#3
provided there isn't extensive overlap in the FR, and the mid-bass driver's HF roll-off is not too hashy, there's no reason a simple cap wouldn't be worth a try - which certainly makes spending a bit more on nice film caps worth considering.
 
#4
Sadly, this is how many super cheap commercial speakers are designed. This gives the minimum protection to the tweeter and you are listening to the woofer's breakup...

But, on the glass half full side of the coin, it IS possible to pull this off in a halfway decent fashion if (a) woofer breakup is benign, and (b) the tweeter can handle the extra power and low frequency content that will still reach it.

Typically this would mean a circa 6" poly cone woofer, and a ferrite magnet tweeter with a 2.5-3.5k Hz resonant frequency and a very low impedance peak.

There will still not be any baffle step correction or any other circuitry to shape the frequency response, remove peaks, etc. But I have to admit the very first speaker I ever built using some drivers from RadioShack had a single cap in series with the tweeter. It was used in a bookshelf, which acted like a huge baffle. I recall that it didn't sound half bad!

-Charlie
 
#5
Hi,

Surely, it can be done, and the results can be rather good - but you need to choose your drivers carefully. Paper or poly midrange with minimal breakup and smooth roll-off is most important. For tweeter, though, I´d suggest 2. order crossover with somewhat smaller cap and larger coil than textbook suggests - experimentation and measuring is the key. 1. order x-over doesn't work very well for tweeters, IMHO. If you choose right drivers you can actually get true acoustic 2. order roll-offs for both mid and tweeter.

I've done it myself with Eminence Alpha 6 mid and Celestion HF-50 tweeter in a 3-way setup - no passive crossover for Alpha (4. order L-R active highpass filter at around 300 Hz); 4,7 uF cap in series and 0,22 mH inductor parallel with the tweeter. Acoustical crossover point is around 5,5...6 kHz. Sounds very good, almost like a good fullranger. Before that I tried different solutions - 1. order for both drivers; mixed 1. order for Alpha - 2. (even 3.) order for Celestion; no crossover for Alpha - 1. and 3. order for Celestion; different values and attenuation resistors... In the end, second simplest solution (2. order with non-textbook values and no padding resistors) worked the best.

The trade-off of this kind of setup is bad vertical polar response - the x-over point is necessarily high, and this shows.
 
#6
technofreak said:
Hi,

Surely, it can be done, and the results can be rather good - but you need to choose your drivers carefully....
Click to expand...

Surely it HAS been done by Burhoe many years ago with the very successful EPI 100 speaker. Just a 10 uF cap in series with the tweeter and that's it. The formula for success is as you pointed out: picking the right drivers that can handle themselvels in that kind of environment.

The circs 1990 Sonus Faber Extrema approx. $12,000 speaker came close with just a resistor in series with the tweeter and parallel coil and a coil in series with the bass/mid.
 
#7
If you haven't bought the drivers, yet, then I'd suggest skip the multi-way configuration and head for a decent wide-ranger.

(Eventually, the cheap solution might probably cost you much more then expected.... )

If you got all the drivers on hand already, then take a look at this site:
Loudspeaker Design Software

I found it easy to use and very accurate if the impedance data is available.

Driver's impedance data can be obtained from measuring tool like WT3. It's a good tool and worth the money. Very useful if you are going to stay in this hobby.
 
#9
thanks to all.
i have already bought following drivers.

Visaton - Lautsprecher und Zubehör, Loudspeakers and Accessories

Visaton - Lautsprecher und Zubehör, Loudspeakers and Accessories

and i am building exactly same speaker as in the following link,
Visaton - Lautsprecher und Zubehör, Loudspeakers and Accessories

AR-SXO seems to be the simplest possible solution for me compared to the one given in the above link.

i have one more question.
what are your thoughts on active xover? i am planning to use LR4 xover ckt as in below link,
Linkwitz-Riley Electronic Crossover

which would give better results? passive or active? or in other words which one would suit more for the drivers i have chosen?

thanks,
Pranam
 
#13
soundnovice said:
I will build as per ckt given by Visaton, but my intention would be to reduce the xover components and cost as much as possible
Click to expand...

Why? From what I read Visatons designs work reasonably well and the components do not look unreasonably expensive.
Your dollar but if you want to roll your own why buy such difficult drivers to work with? Tweeters Fs is too high to work well crossed single order to go with that woofer, with a single cap in series you should go for a XO frequency higher than 5k, my experience says there will be a hole in the integrated response
 
#14
CharlieLaub said:
Sadly, this is how many super cheap commercial speakers are designed. This gives the minimum protection to the tweeter and you are listening to the woofer's breakup...

But, on the glass half full side of the coin, it IS possible to pull this off in a halfway decent fashion if (a) woofer breakup is benign, and (b) the tweeter can handle the extra power and low frequency content that will still reach it.

Typically this would mean a circa 6" poly cone woofer, and a ferrite magnet tweeter with a 2.5-3.5k Hz resonant frequency and a very low impedance peak.

There will still not be any baffle step correction or any other circuitry to shape the frequency response, remove peaks, etc. But I have to admit the very first speaker I ever built using some drivers from RadioShack had a single cap in series with the tweeter. It was used in a bookshelf, which acted like a huge baffle. I recall that it didn't sound half bad!

-Charlie
Click to expand...

How is it sadly? I think it is the way to design cheap speaker. Would it be very clever idea if very cheap speaker had very complex 24 part crossover (that costs at least as much as drivers)? 😉

In theory (and in practice) I like low order crossovers and drivers that have wide usable frequency response. Low order XO makes less phase shift and "it feels" that very steep XO's make speaker sound "wrong", music does not breathe like with fullrange for example.

Does this looks like good tweeter for 1.st order capacitor XO ? Low impedance peak.

MarkAudio has done woofer's that have natural baffle step correction, they might be worth to try, though BSC looks bit too heavy. http://www.hifitalo.com/pdf/EL166-data.pdf
 
#16
You might find this site interesting, site concentrates on passive crossovers.

The site is easy to follow, charts, diagrams and even a calculator. Yes the info is for car setups, but the theories are all the same, IMO.

Zobel networks are also covered.

http://www.the12volt.com/caraudio/cross.asp




soundnovice said:
hi all,
in its simplest form (neglecting sound quality) can i just connect a suitable capacitor in series with tweeter and give power amp output to the drivers directly? which cap is to be used? polar or non-polar

thanks,
Rranam
Click to expand...
 
#20
soundnovice said:
can anybody take a look at the FR curves of the woofer (link posted above) and tell me if it has 'smooth roll-off'? i am not able to make out how the smooth roll-off is supposed to look
Click to expand...

It should look smooth.😀

Yes it looks quite smooth but roll off is at very high frequency and Visaton shows no off axis curves.

Can we believe Visaton measurements like Seas's for example?

This Seas driver has very nice curve, it is almost fullrange.
 
Status
Not open for further replies.
Top Bottom